绝缘子憎水性图像水珠/水迹形状提取算法  

Algorithm of Distilling Information of Hydrophobic Image of Insulators

作　　者：祁兵[1] 唐良瑞[1] 赵春辉[1] 

机构地区：[1]华北电力大学电气与电子工程学院,北京102206

出　　处：《电工技术学报》2008年第6期19-24,共6页Transactions of China Electrotechnical Society

摘　　要：绝缘子表面憎水性检测是判别绝缘子性能优劣的主要手段。传统的检测方法具有其主观性,图像处理技术的引入为憎水性检测提供了更为客观的方法。考虑到绝缘子所处环境的复杂性(污秽、光照等)以及憎水性图像本身的特殊性(水的透明性),本文首先对绝缘子憎水性图像进行自适应直方图均衡和基于模糊逻辑的滤波,以达到增强图像的效果;然后通过基于模糊熵的自动阈值分割方法来获取水珠(水迹)的形状信息,最后用数学形态学的二值重构开运算来去除由于分割引入的噪声。实验统计结果表明,该方法能够对各级憎水性图像进行有效地分割。Hydrophobic measurement is a main way to evaluate the performance of insulators. The usage of image processing technology provides an accurate method of detecting the hydrophobic grades. In this paper, considering the complexity of hydrophobic image, firstly, image intensifying is obtained by self-adaptive histogram equalization and fuzzy de-noising. Secondly, automatic threshold algorithm based on fuzzy entropy is used for image segmentation. Finally, morphology reconstruction is implemented to eliminate noises brought by fuzzy segmentation. Experiments show that this algorithm is universal to all hydrophobic images.

关 键 词：憎水性 自适应直方图均衡 模糊逻辑模糊熵 二值重构 

分 类 号：TM45[电气工程—电器]